[EM] Classifying 3-cand scenarios. LNHarm methods again.

[Kristofer Munsterhjelm]

James Gilmour wrote:
> robert bristow-johnson  > Sent: Monday, April 19, 2010 4:03 AM
>> I dunno about France, but is that the case in Italy?  or Israel?  I
>>  thought there were a bunch of countries with a half dozen
>> contending parties or more.  it looks to me that even the UK has
>> three significant parties.
> 
> It seems to me that in mentioning these particular countries you are
> mixing up two aspects of voting system reform that should always be
> kept completely separate, namely choosing a single-winner voting
> system for single-OFFICE elections and choosing a system for the
> election of representatives to representative assemblies.
> Single-winner voting systems should never be used to elect the 
> members of a representative assembly because, except by chance,
> single-winner voting systems cannot deliver the primary requirement -
> an assembly properly representative of those who voted.
> 
> The reference to France could be to the Presidential election  -
> that is a single-winner election by popular vote, but it uses Top-Two
> Run-Off with occasional disastrous consequences.  The members of the
> French National Assembly are elected from single-seat electoral
> districts, also with two-round run-off, and so that Assembly is not
> properly representative of those who vote.

France has tried proportional representation multiple times, but they've 
reverted to single-seat majoritarian districts every time. I am not sure 
why they did so, but from a cursory glance at web articles, it appears 
that the first period (from the end of WWII to 1956) ended because of 
instability, whereas the second (1986) ended partly because the people 
didn't want a return to that instability and partly because the 
president was from the "wrong" coalition.

> In Italy the national Parliament was from 1945 to 1993 elected by
> closed-list party-list national PR, with two very low thresholds. 
> Italy then flirted with MMP but went back to party-list PR in 2005
> but with a 55% seat distortion to favour the coalition with most 
> votes.  Israel uses closed-list party-list national PR with a very
> low national vote (artificial) threshold.  Both countries have highly
> fragmented party systems  -  perhaps a consequence of using
> closed-list versions of party-list PR voting systems.

Yet for national elections, Norway has both a relatively closed party 
list system (it takes a large fraction of the voters to change the list 
order) and a low threshold (4% for the top-up seats, nothing for the 
rest), but the party politics are not very fragmented. Perhaps the use 
of modified Sainte-Laguë keeps the instability at bay, but that doesn't 
seem likely.

Israel being a single constituency might explain fragmentation there, as 
the number of seats in a constituency forms an implicit threshold (and 
in the Norwegian case, the mechanism that ameliorates this has an 
*explicit* threshold), but as for Italy, I don't know. There might 
simply be a non-technological aspect to which democracies would become 
fragmented under PR and which would not.